Speed of recursive forecasting has gotten worse
mdancho84 opened this issue · 4 comments
It looks like the speed of the recursive() forecasting has gotten worse.
#197 The only thing I can think of is this may have caused a slowdown somewhere but not sure.
Sorry if this is my fault! FWIW, I don't see any performance regression when I run some example models using the CRAN release versus the dev version for recursive models that forecast one row at a time, and I'm still seeing the performance improvements in the cases where the forecasts can be generated in chunks. If you have a specific example of a model that exhibits a performance regression I'd be happy to try to take a look.
I'm taking this one down - Performance looks OK.
I'm so relieved! I was really worried I'd somehow made things worse!